ems_data_clean = read.csv("./data/EMS_Incident_Dispatch_Data.csv") %>%
  select(INITIAL_SEVERITY_LEVEL_CODE, FINAL_SEVERITY_LEVEL_CODE, INITIAL_CALL_TYPE,
         DISPATCH_RESPONSE_SECONDS_QY, INCIDENT_TRAVEL_TM_SECONDS_QY, HELD_INDICATOR, BOROUGH, ZIPCODE, INCIDENT_DISPOSITION_CODE, INCIDENT_DATETIME) %>% 
  janitor::clean_names() %>% 
  separate(col = incident_datetime, into = c('date', 'time'), sep = ' ') %>% 
  separate(col = date, into = c("month","day"), sep = '/') %>% 
  mutate(month = factor(month, levels = c("01", "02", "03", "04", "05", "06", "07", "08", "09", "10", "11", "12"), labels = c("Jan", "Feb", "Mar", "Apr", "May", "June", "Jul", "Aug", "Sept", "Oct", "Nov", "Dec"), ordered = TRUE)) %>% 
  mutate(arrival_outcome = ifelse(incident_disposition_code == "83", "dead", "alive"),
         arrival_outcome = recode_factor(arrival_outcome, `0` = "alive", `1` = "dead"),
         initial_severity_level_code = factor(initial_severity_level_code, 
 levels = c("1", "2","3", "4", "5", "6", "7", "8", "9"), ordered = TRUE),
         final_severity_level_code = factor(final_severity_level_code, 
                                            levels = c("1", "2", "3", "4", "5", "6", "7", "8"), ordered = TRUE), 
         held_indicator = recode(held_indicator, "N" = "no", "Y" = "yes")) %>%
  mutate(neighbourhood = recode(zipcode, "10026" = "central harlem", "10027" = "central harlem", "10030" = "central harlem", "10037" = "central harlem", "10039" = "central harlem", "10001" = "chelsea and clinton", "10001" = "chelsea and clinton", "10011" = "chelsea and clinton", "10018" = "chelsea and clinton", "10019" = "chelsea and clinton", "10020" = "chelsea and clinton", "10036" = "chelsea and clinton",  "10029" = "east harlem", "10035" = "east harlem", "10010" = "gramercy park and murray hill", "10016" = "gramercy park and murray hill", "10017" = "gramercy park and murray hill", "10022" = "gramercy park and murray hill", "10012" = "greenwich village and soho", "10013" = "greenwich village and soho", "10014" = "greenwich village and soho", "10004" = "lower manhattan", "10005" = "lower manhattan", "10006" = "lower manhattan", "10007" = "lower manhattan", "10038" = "lower manhattan", "10280" = "lower manhattan", "10002" = "lower east side", "10003" = "lower east side", "10009" = "lower east side", "10021" = "upper east side", "10028" = "upper east side", "10044" = "upper east side", "10065" = "upper east side", "10075" = "upper east side", "10128" = "upper east side", "10023" = "upper west side", "10024" = "upper west side", "10025" = "upper west side", "10031" = "inwood and washington heights", "10032" = "inwood and washington heights", "10033" = "inwood and washington heights", "10034" = "inwood and washington heights", "10040" = "inwood and washington heights" )
  ) %>%  
  drop_na(neighbourhood) %>% 
    select(-incident_disposition_code)

```

Map:

graph_data = ems_data_clean %>% 
  mutate(zipcode = as.character(zipcode))

# count the number of death's upon arrival by zip code
zip_count = graph_data %>% 
  select(arrival_outcome, zipcode, neighbourhood) %>%
  group_by(arrival_outcome, zipcode, neighbourhood) %>% 
  drop_na(arrival_outcome) %>% 
  summarise_(n_death = ~n()) %>%
  filter(arrival_outcome == "dead") 
zip_count
## # A tibble: 41 x 4
## # Groups:   arrival_outcome, zipcode [42]
##    arrival_outcome zipcode neighbourhood                 n_death
##    <fct>           <chr>   <chr>                           <int>
##  1 dead            10001   chelsea and clinton                35
##  2 dead            10002   lower east side                   140
##  3 dead            10003   lower east side                    54
##  4 dead            10004   lower manhattan                     5
##  5 dead            10005   lower manhattan                     6
##  6 dead            10006   lower manhattan                     2
##  7 dead            10007   lower manhattan                     2
##  8 dead            10009   lower east side                    81
##  9 dead            10010   gramercy park and murray hill      36
## 10 dead            10011   chelsea and clinton                56
## # … with 31 more rows
#import shape file for NYC zipcodes 
city_map = readOGR(dsn = './data/ZIP_CODE_040114/ZIP_CODE_040114.shp', encoding = "UTF-8")
## OGR data source with driver: ESRI Shapefile 
## Source: "/Users/Jai/Desktop/mph3/data science I/jacab_p8105final/data/ZIP_CODE_040114/ZIP_CODE_040114.shp", layer: "ZIP_CODE_040114"
## with 263 features
## It has 12 fields
city_map@data = left_join(city_map@data, zip_count, by = c("ZIPCODE" = "zipcode")) 



city_map_transform = spTransform(city_map, CRS("+init=epsg:4326"))


#popups for polygons
label_popup = paste0(
  "<strong>Zipcode: </strong>",
  city_map$ZIPCODE,
  "<br><strong>Number of Deaths: </strong>",
  city_map$n_death
)



# set bins
death_bins = c(0, 10, 20, 30, 40, 50, 80, 110, 140, 166)

# set pals
death_pal = colorBin('Reds', bins = death_bins, na.color = 'white')

# choropleth map for # of deaths per zipcode
leaflet(data = city_map_transform) %>% 
  addProviderTiles('CartoDB.Positron') %>% 
  addPolygons(fillColor = ~death_pal(n_death),
              fillOpacity = 0.9,
              color = "grey",
              weight = 1,
              popup = label_popup,
              highlightOptions = highlightOptions(color = "black", weight = 2,
                                                  bringToFront = TRUE)) %>% 
  addLegend('bottomleft',
            pal = death_pal,
            values = ~n_death,
            title = 'Number of Dead on Arrival by Zipcode',
            opacity = 1)